Histopathologic Changes of the Inner ear in Rhesus Monkeys After Intratympanic Gentamicin Injection and Vestibular Prosthesis Electrode Array Implantation

Abstract

Bilateral vestibular deficiency (BVD) due to gentamicin ototoxicity can significantly impact quality of life and result in large socioeconomic burdens. Restoring sensation of head rotation using an implantable multichannel vestibular prosthesis (MVP) is a promising treatment approach that has been tested in animals and humans. However, uncertainty remains regarding the histopathologic effects of gentamicin ototoxicity alone or in combination with electrode implantation. Understanding these histological changes is important because selective MVP-driven stimulation of semicircular canals (SCCs) depends on persistence of primary afferent innervation in each SCC crista despite both the primary cause of BVD (e.g., ototoxic injury) and surgical trauma associated with MVP implantation. Retraction of primary afferents out of the cristae and back toward Scarpa’s ganglion would render spatially selective stimulation difficult to achieve and could limit utility of an MVP that relies on electrodes implanted in the lumen of each ampulla. We investigated histopathologic changes of the inner ear associated with intratympanic gentamicin (ITG) injection and/or MVP electrode array implantation in 11 temporal bones from six rhesus macaque monkeys. Hematoxylin and eosin-stained 10-μm temporal bone sections were examined under light microscopy for four treatment groups: normal (three ears), ITG-only (two ears), MVP-only (two ears), and ITG + MVP (four ears). We estimated vestibular hair cell (HC) surface densities for each sensory neuroepithelium and compared findings across end organs and treatment groups. In ITG-only, MVP-only, and ITG + MVP ears, we observed decreased but persistent ampullary nerve fibers of SCC cristae despite ITG treatment and/or MVP electrode implantation. ITG-only and ITG + MVP ears exhibited neuroepithelial thinning and loss of type I HCs in the cristae but little effect on the maculae. MVP-only and ITG + MVP ears exhibited no signs of trauma to the cochlea or otolith end organs except in a single case of saccular injury due to over-insertion of the posterior SCC electrode. While implanted electrodes reached to within 50–760 μm of the target cristae and were usually ensheathed in a thin fibrotic capsule, dense fibrotic reaction and osteoneogenesis were each observed in only one of six electrode tracts examined. Consistent with physiologic studies that have demonstrated directionally appropriate vestibulo-ocular reflex responses to MVP electrical stimulation years after implantation in these animals, histologic findings in the present study indicate that although intralabyrinthine MVP implantation causes some inner ear trauma, it can be accomplished without destroying the distal afferent fibers an MVP is designed to excite.

Appendix

Fig. 7

Electrode array placement. A Posterolateral view of 3D CT surface reconstruction showing electrode array leads implanted in the left labyrinth of a rhesus monkey via the mastoid cavity. a lead to anterior and horizontal ampullae; b lead to posterior ampulla; c common crus reference electrode; d neck reference electrode; M mandibular ramus; Z zygomatic arch; ANT, POST, SUP, INF anterior, posterior, superior, inferior. B Oblique CT cut through the plane of the basal turn of the cochlea [Co], showing bifurcated electrode array [a] entering the ampullae of the superior [s-scc] and horizontal [h-scc] semicircular canals. Part of the neck reference electrode [d] is also visible, but the posterior SCC electrode array is not included in this section. AM, PL anteromedial, posterolateral. C Image of forked electrode arrays for h-scc and s-scc ampullas (1), posterior SCC (2), and common crus reference (inset) electrodes. A and B reproduced with permission from Dai et al. 2011b.